This study was an observational case-control study on 23 preterm neonates with NEC (NEC group), and 30 preterm neonates without NEC (control group) at a tertiary referral neonatal intensive care unit between July 2014 and October 2016. The Institutional Review Board of Seoul St. Mary’s Hospital approved this study and waived to obtain informed consent.

NEC was diagnosed on the basis of the patient’s systemic, intestinal, and radiologic signs and staged according to the modified Bell’s classification [1,16]. Preterm neonates who had been managed under the diagnosis of NEC (NEC stage ≥1) and had undergone a comprehensive echocardiographic evluation before 14 days of age and within 7 days before the diagnosis of NEC comprised NEC group (n=23). The control group (n=30) consisted of preterm neonates who received neonatal intensive care during the same period but did not have NEC and echocardiographic study before 14 days of age approximated to the day of life of their NEC counterparts. Neonates who had structural heart diseases, structural anomalies in the gastrointestinal tract or culture-proven sepsis were excluded from the study. Basic clinical characteristics and echocardiographic hemodynamic data were collected from our institutional database, and compared between the NEC and control groups. The NEC group was further classified into three subgroups according to the NEC stages, and echocardiographic hemodynamic data were compared among the subgroups.

Echocardiographic hemodynamic assessments were done by an experienced pediatric cardiologist according to the protocol of our pediatric cardiology program, and all images were checked for appropriate quality required for analyses by an experienced consultant pediatric cardiologist, as previously reported [17]. In summary, for conventional assessment, the left ventricular ejection fraction (LVEF), and left ventricular fractional shortening (LVFS) were measured from the parasternal long-axis view. From the spectral pulsed-wave Doppler of the DA, the resistive index (RI), a surrogate of peripheral vascular resistance, was calculated, as follows: RI=(peak systolic velocity–lowest diastolic velocity)/peak systolic velocity (Figure 1) [18]. The DABF was derived by multiplying, the cross-sectional area of the DA, the velocity time integral (VTI) of the DA Doppler at the level of the diaphragm in a low subcostal sagittal view, and the heart rate [17]. The diastolic reverse flow of the DA, if any, was calculated by multiplying the VTI of diastolic reverse flow, the cross-sectional area of the DA, and the heart rate (Figure 1). The ratio of the diastolic reverse flow of DA to the DABF was also derived. To obtain pulsed-wave Doppler images of the DA for calculating the DABF and RI, an appropriate Doppler angle correction was applied. In case of ductus arteriosus, its diameter and transductal Doppler velocity were measured in a high parasternal ‘ductal’ view. The dimensions of left atrium and aortic root were measured in the parasternal long-axis view. An echocardiographic hemodynamically significant ductus arteriosus (HSDA) was defined as the presence of ductus arteriosus ≥1.5 mm with an unrestrictive pulsatile transductal flow (peak Doppler velocity <2.0 m/sec) and a left atrium to aortic ratio of ≥1.5. All echocardiographic flow indices were recorded from the average value of three consecutive cardiac cycles.

Data were analyzed using commercially available software (PASW statistics version 18 for Window; IBM, Armonk, NY, USA). Data were expressed as mean±standard deviation, median (interquartile range [IQR]), or number (%). Comparisons of basic clinical characteristics and echocardiographic parameters between the NEC and control groups were conducted using a 2-tailed, independent-samples t-test for normally distributed data and a 2-tailed Mann-Whitney U-test for data that were not normally distributed, or Fisher exact test when appropriate. Comparison of echocardiographic parameters among the three NEC subgroups and the control group were performed using the Kruskal-Wallis test. Univariate and multivariate logistic regression analyses were done to identify the risk factors for NEC development. A P-value less than 0.05 was considered statistically significant.

A total of 53 preterm neonates were enrolled, including 23 neonates who developed NEC (NEC group) and 30 neonates without NEC (control group) during the study period. The basic clinical characteristics of each group are summarized in Table 1. The two groups had similar basic clinical characteristics with regard to gestational age, birth weight, Apgar score, mode of delivery, mode of feeding, types of central venous catheters and the ratio and mode of invasive ventilator care. In the NEC group, NEC was diagnosed at a median of 8 days (IQR, 5 to 11) after birth. There were 12 neonates (52%) with stage 1 NEC, five neonates (22%) with stage 2 NEC, and six neonates (26%) with stage 3 NEC according to the modified Bell’s classification. Echocardiographic evaluation were done at a median 5 days (IQR, 3 to 9) for the NEC group, and 4 days (IQR, 3 to 8) for the control group after birth, respectively (P=not significant). In the NEC group, echocardiography was done at a median 2 (IQR, 1 to 2.5) before the diagnosis of NEC.

The echocardiographic hemodynamic characteristics of each group are summarized in Table 2. Compared with the control group, the NEC group had significantly decreased Doppler diastolic velocity, and significantly increased RI, diastolic reverse flow of the DA and the ratio of diastolic reverse flow of the DA to the DABF. Other parameters including blood pressure, LVEF, LVFS, and peak systolic velocity of the DA were similar between the groups. The incidence of ductus arteriosus, its diameter, left atrium to aortic ratio, and incidence of HSDA were also similar between the groups.

The subgroup analysis in the NEC group according to the NEC stages showed significant differences in the peak systolic velocity, lowest diastolic velocity, RI, DABF, diastolic reverse flow of the DA and the ratio of diastolic reverse flow of the DA to DABF among the subgroups (Table 3). The higher the NEC stage, the more significant decrease in the peak systolic velocity, lowest diastolic velocity and DABF, whereas more significant increase in the RI, diastolic reverse flow of DA, and the ratio of diastolic reverse flow of DA to DABF were noted. The proportions of patients with HSDA were similar among NEC subgroups.

To identify the risk factors for the development of NEC, logistic regression analyses were done for echocardiographic hemodynamic parameters (Table 4). In univariate analysis, the lowest diastolic velocity of the DA, RI, diastolic reverse flow of the DA and the ratio of diastolic reverse flow of the DA to the DABF were associated with the development of NEC. In multivariate analysis, the increasing RI of DA flow was an independent risk factor for development of NEC (P=0.008).